Thanks Tom,
I'll give it a try, the whole setup is synthetic anyway, should not be too
much of a trouble

Best regards
Seref


On Sun, Mar 15, 2015 at 3:16 PM, Tom Lane <t...@sss.pgh.pa.us> wrote:

> Seref Arikan <serefari...@kurumsalteknoloji.com> writes:
> > Another idea that popped into my head is to consider table partitioning.
> Am
> > I correct to assume that (inevitable) joins could benefit from partitions
> > when sub query criteria constraints results to a particular partition?
>
> That's too vague to answer really.  My suspicion is that you'd still end
> up having to modify the queries, ie hand-factoring the joins, but maybe
> you wouldn't need to.  You could experiment easily enough; just set up a
> toy database with not much data in it and see if you get plans that don't
> scan all the partitions.
>
>                         regards, tom lane
>

Reply via email to